Diabetes affects between 6.5 and 13 million Americans and is associated with an expenditure of over 44 billion in direct health care costs per year, Depression occurs as a comorbid condition in 23% of European Americans with diabetes, and the rate of depression among ethnic minorities is considerably higher. The prevalence of sub clinical depression increases the impact even more. Depression is associated with decreased self-management, poor glycemic control, and increased risk for complications. Explanatory models, based almost exclusively on cross-sectional research, suggest that depressive affect has both direct and indirect effects on glycemic control through self-management, and that glycemic control can affect depressive affect directly. These relationships vary considerably from patient to patient and are affected by such potentially malleable variables as psychological resources, family relationships, and chronic stress. Given the increased behavioral and biological risks that depressive affect pose for these patients, longitudinal studies with implications for intervention are needed to describe the ways in which these constructs influence each other over time. Using a "micro longitudinal" approach with a community based sample that employs daily data collection via telephone interview, we propose to examine the day-to-day variation among depressive affect, disease-management and glycemic control over 21 days in patients with diabetes. Our aims are (1) to provide greater clarity and specificity to the sequence of causative effects over time; (2) to identify those patients factors that qualify these relationships in an effort to account for the widely documented variability found among these patients; and (3) to identify spouse/partner-related factors that influence these processes to address the social context in which these patients manage diabetes. The proposed longitudinal research has major implications for intervention: it will provide a more complete description of the causative effects of the full range of depressive affect, disease-management and glycemic control over time; and it will identify variations among sub groups of patients that may warrant specific interventions.